Content
57%Weight 40%Scale 1-5Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.
The body is a well-organized CLI reference with genuinely executable commands and useful agent-oriented guidance, held back by duplication, inconsistent command-group naming, and the absence of an explicit validated workflow for the document edit/export cycle. Command tables that would serve better as a one-level-deep reference file are inlined in SKILL.md. Strongest on actionability, weakest on workflow structure and file organization.
Suggestions
Add an explicit end-to-end workflow with validation checkpoints (e.g. 1. `document import input.docx -o work.json` 2. edit via Writer/Calc/Impress commands 3. `document save` 4. `export render output.pdf` 5. verify the output file exists and is non-empty before reporting success).
Move the full command-group tables to a `references/COMMANDS.md` file and keep only a quick-start example in SKILL.md, making the reference one level deep.
Reconcile command-group naming with the actual CLI invocations (the "Document" table is invoked as `project`/`document` and the "Export Group" as `export` in examples), and consolidate the duplicated REPL and `project new` examples.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body is mostly efficient — command tables and code blocks with little explanation of concepts Claude already knows — but includes unnecessary duplication (REPL mode described in both Usage and Examples, `project new` shown twice) and full command-reference tables inlined that could be tightened or split out. It is above anchor 2 (no heavily padded concept explanations) but not anchor 4 (the repetition and inline reference bulk could be trimmed). | 3 / 5 |
Actionability | Mostly executable guidance: concrete invocations like `cli-anything-libreoffice --json project info -p project.json`, `document import input.docx -o work.json`, and `export render output.pdf --overwrite`. Minor gaps keep it below 5: table commands lack argument examples, and group naming is inconsistent (the "Document" group is invoked as `project`, the "Export Group" as `export`). | 4 / 5 |
Workflow Clarity | The "For AI Agents" numbered practices and examples imply a sequence (create/import → edit → export → verify), but there is no explicit end-to-end workflow with validation checkpoints. Document manipulation is a feedback-loop-critical context and validation appears only as a single bullet ("Verify outputs exist"), so workflow clarity is capped at 3 per the rubric guidelines. | 3 / 5 |
Progressive Disclosure | Sections are clearly organized with headers, but the ~230-line body inlines full command-group reference tables that belong in a separate reference file, and its only pointers (README.md, TEST.md, HARNESS.md "in the package") are not bundled files in the skill directory. This matches "some structure but content that should be separate is inline"; it is above anchor 2 because structure and navigation are present and clear. | 3 / 5 |
Total | 13 / 20 Passed |